Project Summary/Abstract: Latently-infected CD4+ T cells are thought to be the main barrier to HIV eradication or functional cure, and viral reactivation from these cells likely contributes to the organ inflammation and damage observed on antiretroviral therapy. Major impediments to the development of more effective latency-reversing agents (LRAs) include the lack of knowledge about the mechanisms that govern latency and latency reversal in vivo, and the degree to which these are recapitulated by latency models in vitro. The lack of agreement between latency models and the incomplete success of human trials with LRAs suggest that it is critical to understand why different LRAs do or do not work in vivo. We have developed a new ?transcription profiling? approach that can simultaneously measure the degree to which different mechanisms contribute to reversible inhibition of HIV transcription in vivo. By applying this approach to cells from ART-suppressed patients, we have generated preliminary data suggesting a new paradigm in which latency is not (as commonly assumed) due to a block to HIV transcriptional initiation, the block to proximal elongation is greater and more pervasive than previously realized, and the main reversible blocks to HIV transcription are a previously-unrecognized block to distal transcription/polyadenylation (completion) and a block to multiple-splicing. In addition, we have intriguing new data suggesting that LRAs may act selectively on the different mechanistic blocks to HIV transcription. This study will utilize samples from clinical trials of humans treated with LRAs (aims 1 and 2) to better understand how they reverse the mechanisms of latency in vivo and to identify the optimum model to test new agents in vitro (aim 3). In aim 1, we will apply our transcription profiling approach to samples from humans treated with disulfiram, vorinostat, panobinostat, and romidepsin. We hypothesize that these agents preferentially increase HIV transcriptional initiation and elongation but have less ability to overcome blocks to completion and splicing. In aim 2, we will apply our approach to blood and gut samples from clinical trials of humans treated with agonists of toll-like receptor (TLR) 7 and 9 to understand how these agents reverse latency and lead to death of infected cells in vivo. We hypothesize that TLR agonists can overcome later blocks to HIV transcription, increasing the completed transcripts (and HIV protein/antigen) that may facilitate clearance by intrinsic cell defenses or immune killing. In aim 3, we will compare in vitro models of latency based on the degree to which they recapitulate in vivo mechanisms of latency and responses to LRAs. We will then select the best model and test new combinations of agents for their ability to increase completed/spliced transcripts and lead to death of infected cells. The results from these 3 aims should provide critical new insights on the degree to which existing LRAs reverse the different mechanisms of latency in vivo (aims 1 and 2), the effects that correlate with clearance of reactivated cells (aims 2 and 3), the best system to study LRAs in the laboratory (aim 3), and new combinations that can lead to more effective latency reversal and/or killing of infected cells (aim 3).